DESCRIPTION: Ordinary conversational speech arrives very rapidly, typically averaging between 140 to 180 words per minute. From this acoustically complex and often poorly articulated speech stream the listener must rapidly segment the signal into words, extract the linguistic structure and semantic content, and then internally organize this information for comprehension and later recall. The objective of this proposal is to study rapid speech processing and memory in elderly adults. At the theoretical level, the investigators wish to understand how spoken language comprehension and memory operate within known age-related changes in auditory processing efficiency and transient memory capacity. The program has three interleaved components. The first will be to use the technique of self-paced listening to investigate elderly adults' spontaneous resource allocation strategies while listening to speech. This component will include the use of speech rate manipulations to further study the interaction between speech content, and age-related changes in processing speed, on speech comprehension and memory. The second component will examine the ways in which young and elderly adults use the intonation and stress patterns of natural speech to facilitate the structural analysis and comprehension of the speech input. The final component will use a technique of whole-word gating to explore the nature of the boundary conditions that operate on the effective use of linguistic context in word recognition in meaningful speech. Underlying this research is a question of major theoretical importance, and practical significance. This question is the degree to which time-allocation strategies, and the use of linguistic context in speech processing, represent flexibly deployable operations or whether they represent fixed modes of functioning in the aging cognitive system.